Remote diagnosis method for ground fault of beam-pumped well

Abstract

The invention belongs to the technical field of oil pumping unit fault diagnosis, and particularly relates to a remote diagnosis method for ground faults of a beam-pumping unit well. According to the method, the corresponding curves used for judging the fault type of the oil pumping unit are obtained, and whether the oil pumping unit breaks down or not and the fault type are judged according to all the obtained curves; the curve for judging the fault type of the oil pumping unit only comprises a current horsehead displacement curve and a current oil pumping unit well indicator diagram, or only comprises a current horsehead load curve and a current oil pumping unit well indicator diagram, or comprises the current horsehead displacement curve, the current horsehead load curve and the current oil pumping unit well indicator diagram; through combined analysis of all curves, the problems that ground fault detection of the oil pumping unit depends on manual inspection and fault detection is not timely are solved.

Description

Technical Field

[0001] This invention belongs to the field of oil pumping unit fault diagnosis technology, specifically relating to a remote diagnosis method for surface faults in beam pumping unit wells. Background Technology

[0002] my country has over 200,000 oil pumping units of various types, with beam pumping units accounting for more than 90% of the total in operation in oil fields. During field operation, beam pumping units frequently experience vibrations, shaking, abnormal noises, and even the beam falling and damaging the wellhead due to beam head cracking. Failure to detect these issues promptly during inspections and take appropriate measures can adversely affect the stable operation of oil and gas production. Currently, inspections mainly rely on manual on-site inspections, video inspections, and drone inspections. These methods all suffer from varying degrees of problems, including high costs, low efficiency, and delayed fault detection.

[0003] A Chinese invention patent with authorization announcement number CN110163302B discloses a dynamometer card recognition method based on a regularized attention convolutional neural network. This method uses time-series data of displacement and load of the pumping unit's pumping head movement to draw the contour curve of the dynamometer card and trains and tests the constructed dynamometer card recognition model. Real-time collected pumping unit operating data is input into the trained dynamometer card recognition model for recognition. If a fault occurs, the diagnostic results are sent to the management personnel via email and SMS. Although this method can solve the problem of relying on manual on-site inspection, it can only solve the detection of faults downhole. Faults occurring on the surface (such as pumping unit pumping head cracks and high back pressure) still require manual inspection and cannot detect faults in a timely manner. Summary of the Invention

[0004] The purpose of this invention is to provide a remote diagnostic method for surface faults in beam pumping units, in order to solve the problems of relying on manual inspections and untimely fault detection in pumping unit surface fault detection.

[0005] To address the aforementioned technical problems, this invention provides a remote diagnostic method for surface faults in beam pumping unit wells, the method comprising:

[0006] Obtain the curves corresponding to the type of pumping unit failure, and determine whether the pumping unit has failed and the type of failure based on all the obtained curves; the curves include only the current donkey head displacement curve and the current pumping unit well dynamometer diagram, or only the current donkey head load curve and the current pumping unit well dynamometer diagram, or the current donkey head displacement curve, the current donkey head load curve and the current pumping unit well dynamometer diagram.

[0007] Furthermore, the fault type includes donkey head fracture; the curves corresponding to determining whether the pumping unit has experienced a donkey head fracture fault include the current donkey head displacement curve, the current donkey head load curve, and the current pumping unit well dynamometer diagram; if each curve satisfies all the judgment conditions, then the pumping unit is determined to have experienced a donkey head fracture fault.

[0008] The determination conditions include: dense spikes or abrupt changes appear in the downstroke of the current donkey head displacement curve; near the transition from the downstroke to the upstroke of the current donkey head load curve, the load values ​​of N3 consecutive points are less than the load values ​​of the corresponding normal donkey head load curve, and the degree of less is greater than a third set threshold; near the transition from the downstroke to the upstroke of the current pumping unit well indicator, the load value of a certain point is less than the load value of the corresponding normal pumping unit well indicator, and the degree of less is greater than a fourth set threshold.

[0009] Furthermore, if N2 out of N1 consecutive points in the downstroke of the current donkey head displacement curve have a deviation greater than the first set threshold, then dense burrs are determined to have occurred; the deviation is the ratio of the absolute value obtained by subtracting the measured value of the point from the average displacement of the points before and after the point to the total displacement value of the current stroke, and N1 > N2 ≥ 3.

[0010] Furthermore, if the deviation at a certain point in the downstroke of the current donkey head displacement curve is greater than the second set threshold, a sudden change is determined to have occurred; the deviation is the ratio of the absolute value obtained by subtracting the measured value of the point from the average displacement of the points before and after that point to the total displacement value of the current stroke.

[0011] Furthermore, if it is determined that the pumping unit has a donkey head crack fault, then when the load value at a certain point is less than the load value of the normal pumping unit well indicator diagram, the greater the degree of less than the fourth set degree threshold, the larger the crack that has caused the donkey head crack fault.

[0012] Furthermore, if the displacement of the donkey head during the upward stroke of the current displacement curve decreases or remains unchanged over time, an alarm will be issued.

[0013] Furthermore, if the displacement of the donkey head during the downstroke of the current displacement curve increases or remains unchanged over time, an alarm will be issued.

[0014] Furthermore, if the minimum value of the current donkey head load curve is less than the minimum value of the donkey head load curve of a certain stroke within the previous X strokes, and the degree of less than is greater than the fifth set threshold, then an alarm is triggered.

[0015] Furthermore, if the minimum value of the current donkey load curve is less than the minimum value of a certain donkey load curve in the previous T time periods, and the degree of less than is greater than the sixth set threshold, then an alarm will be triggered.

[0016] Furthermore, the normal donkey head load curve is the theoretical donkey head load curve of the pumping unit or the donkey head load curve when the pumping unit is operating normally; the normal pumping unit well dynamometer diagram is the theoretical pumping unit well dynamometer diagram or the pumping unit well dynamometer diagram when the pumping unit is operating normally.

[0017] Its beneficial effects are as follows: This invention is a pioneering invention. It addresses the fact that existing methods for detecting pumping unit faults primarily target downhole faults, while surface fault detection still relies on manual inspections. Furthermore, when a pumping unit malfunctions, the corresponding operational curves will exhibit different changes. Therefore, this invention utilizes the donkey head displacement curve, donkey head load curve, and existing dynamometer diagrams for downhole fault detection to detect surface faults. This invention obtains the curves corresponding to pumping unit faults and analyzes these curves to determine whether a fault has occurred, thereby identifying the type of fault. This solves the problems of reliance on manual inspections and untimely fault detection in surface pumping unit fault detection. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of a four-bar pumping unit according to an embodiment of the present invention;

[0019] Figure 2 This is the theoretical pumping unit donkey head displacement curve according to an embodiment of the present invention;

[0020] Figure 3 This is the theoretical pumping unit donkey head load curve of an embodiment of the present invention;

[0021] Figure 4 This is a theoretical pumping well dynamometer diagram according to an embodiment of the present invention;

[0022] Figure 5 This is a graph showing the high back pressure fault of the oil pumping unit in this embodiment of the invention.

[0023] Figure 6 This is a graph showing the failure of the oil pumping unit in an embodiment of the present invention when the pumping head cracks.

[0024] Figure 7 These are pumping unit well dynamometer diagrams for different oil wells in this embodiment of the invention when the pumping unit experiences a donkey-head fracture failure.

[0025] Figure 8 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0026] This invention obtains curves corresponding to the types of pumping unit failures, and determines whether a pumping unit has failed and the type of failure based on all the obtained curves. The curves used to determine the type of pumping unit failure include only the current donkey head displacement curve and the current pumping unit well dynamometer diagram, or only the current donkey head load curve and the current pumping unit well dynamometer diagram, or the current donkey head displacement curve, the current donkey head load curve, and the current pumping unit well dynamometer diagram. By combining and analyzing the various curves, the problem of relying on manual inspections and untimely failure detection for pumping unit surface failure detection is solved.

[0027] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0028] Method Example 1:

[0029] This embodiment presents a remote diagnostic method for surface faults in a beam pumping unit well, such as... Figure 8 As shown, based on the analysis of graphical and numerical changes in the beam pumping unit's head displacement curve, head load curve, and well dynamometer diagram, remote inspection and diagnosis of surface faults in beam pumping unit wells can be achieved. The specific generation process of each curve is as follows:

[0030] (1) Donkey head displacement curve.

[0031] like Figure 1 As shown, with AB2 as the zero point of the crank angle, the pumping unit crank rotates counterclockwise at an angular velocity ω. The movement of point B on the crank changes with time t and the crank angle, exhibiting a sinusoidal characteristic. Point C of the walking beam oscillates under the influence of the crank-connecting rod. Affected by the extreme position angle α and the characteristics of the four-bar linkage, the pumping unit moves upward (i.e., the upstroke) within the crank angle θ range of 0 to (π+α), and moves downward (i.e., the downstroke) within the crank angle θ range of (π+α) to 2π. The relationship between the stroke speed ratio coefficient k and the extreme position angle α is as follows:

[0032]

[0033] Under the influence of the quick-return characteristic of the stroke ratio k, the walking beam inclination angle ψ and the donkey head displacement S vary with time t and crank angle θ, approximating a non-standard sinusoidal characteristic. Crank angle θ = ωt, where ω is the crank angular velocity and t is time. Since the motion trajectory of the four-bar linkage can be determined through the constraints and geometric relationships between the links, the crank angle θ determines the corresponding walking beam inclination angle, and the walking beam inclination angle has a linear relationship with the donkey head displacement.

[0034]

[0035] Where S is the donkey head displacement, ψ is the inclination angle of the walking beam, and r is the length of the walking beam's forearm.

[0036] In actual operation, tilt sensors are installed on the walking beam. Each rotation of the crank causes the walking beam to move up and down once. The tilt sensors collect N sets of tilt change data at the same time, and calculate and generate a real-time walking beam displacement curve; for example... Figure 2 As shown, the relationship between the walking beam inclination angle and the donkey head displacement changes with time, exhibiting a sinusoidal characteristic. During normal operation of the pumping unit, this donkey head displacement curve is very smooth. If the pumping unit experiences a ground fault (such as donkey head breakage, non-uniform crank rotation, etc.), the donkey head displacement curve will undergo abrupt changes and produce burrs.

[0037] (2) Donkey head load curve.

[0038] When the pumping unit is moving upwards, the load on the pumping head mainly includes the weight of the rod and the weight of the liquid column above the piston. When the pumping unit is moving downwards, the load on the pumping head mainly consists of the weight of the rod within the liquid column. For example... Figure 3 As shown, during normal operation of the pumping unit, the upward and downward load curves of the donkey head are two parallel horizontal lines. When switching from upward to downward and from downward to upward, the change in the elastic load of the rod column is superimposed on the loads of the upward and downward suspension points of the donkey head. Therefore, the load curves for these transitions are two oblique lines connecting the upward and downward horizontal lines. Thus, during normal operation of the pumping unit, the donkey head load curve is a trapezoid. The slopes of these two oblique lines are related to the elastic deformation δ that the donkey head must overcome before displacement, the crank angular velocity ω, and the time t. The formula for calculating the elastic deformation that the donkey head must overcome before displacement is as follows:

[0039]

[0040] Where F represents the tensile force on the sucker rod string, in Newtons (N); L represents the length of the sucker rod string, in meters (m); E represents the elastic modulus of the sucker rod string, in Pascals (Pa); and A represents the cross-sectional area of ​​the sucker rod string, in m². 2 ); δ is the elastic deformation that the donkey's head must overcome before it undergoes displacement, and its unit is meters (m).

[0041] In actual operation, stress sensors are installed on the walking beam. Each rotation of the crank causes the lifting head to move up and down once. Similar to the tilt sensor, the stress sensors synchronously collect N sets of stress change data to generate a real-time lifting head load curve, which is trapezoidal. If a ground fault occurs in the pumping unit (such as lifting head breakage or non-uniform crank rotation), the lifting head load curve will abruptly change, producing burrs.

[0042] (3) Pumping well dynamometer diagram.

[0043] like Figure 4As shown, the dynamometer diagram in this embodiment only considers the static load borne by the donkey head suspension point, i.e., the elastic deformation of the tubing caused by the static load. The load changes with displacement in a parallelogram shape. In actual operation, the shape of the dynamometer diagram varies from well to well. However, during normal operation of a four-link pumping unit, the dynamometer diagram is usually unchanged between adjacent strokes. If the dynamometer diagram shows a gradual change or abrupt change, the specific cause can be found by comparing the displacement curve and the load curve. For example, if a return stroke or twisting occurs on the dynamometer diagram, it is due to the displacement curve and is mostly a fault in the pumping unit itself. If a large change in load occurs on the dynamometer diagram, it is necessary to compare the load curve and check the site to confirm whether the problem is in the wellbore or on the surface pumping unit. If a needle line or a concave notch appears on the dynamometer diagram at the bottom dead center (i.e., when transitioning from the downstroke to the upstroke), it is necessary to compare the load curve, displacement curve, and check the site to confirm that the donkey head of the pumping unit has cracked.

[0044] The data collected on-site, including the donkey head displacement curve, donkey head load curve, and dynamometer diagram of the pumping unit, is sent to the data receiving device. The host computer then feeds back the relevant parameters received by the data receiving device to the internal digital and graphical displays. It also makes alarms through calculation, comparison, and decision-making. Based on the results of the calculation, comparison, and decision-making, it identifies the type of fault that has occurred in the pumping unit, thus realizing remote diagnosis of surface faults in beam pumping unit wells.

[0045] The main ways to determine the type of fault based on the corresponding curves are as follows: based on the donkey head displacement curve and the pumping unit well dynamometer diagram; based on the donkey head load curve and the pumping unit well dynamometer diagram; based on the donkey head displacement curve, the donkey head load curve and the pumping unit well dynamometer diagram.

[0046] For example, the type of fault occurring in the pumping unit can be determined based on the current donkey head load curve and the current pumping unit well indicator diagram. Figure 5 As shown, compared to a normal donkey-head load curve, the top and bottom of the trapezoidal wave of this donkey-head load curve are narrower, and the two diagonal lines are gentler, longer, and higher. Furthermore, compared to a normal pumping unit well indicator diagram, this pumping unit well indicator diagram appears to be both higher and narrower; therefore, it is determined that the pumping unit is experiencing a high back pressure fault. In this embodiment... Figure 5 , Figure 6 The gray curves in the figures represent the donkey head load curves, with time on the x-axis and load magnitude on the y-axis; the green curves in each figure represent the donkey head displacement curves, with time on the x-axis and displacement magnitude on the y-axis; the red curves in each figure represent the pumping unit well dynamometer diagrams, with displacement magnitude on the x-axis and load magnitude on the y-axis. Note that the actual measurement process was affected by the sensor being installed in the wrong direction. Figure 5 , Figure 6 In the diagram, the displacement curve of the donkey head and the dynamometer diagram are opposite in time stroke to the corresponding time stroke of the load curve of the donkey head.

[0047] In this embodiment, the normal donkey head load curve is the theoretical donkey head load curve of the pumping unit or the donkey head load curve when the pumping unit is running normally, and the normal pumping unit well dynamometer diagram is the theoretical pumping unit well dynamometer diagram or the pumping unit well dynamometer diagram when the pumping unit is running normally.

[0048] Method Example 2:

[0049] This embodiment provides a remote diagnostic method for surface faults in a beam pumping unit well. It determines whether a fault has occurred in the pumping unit and, if so, the type of fault. The generation process for each curve is the same as that described in Method Embodiment 1.

[0050] During one stroke of the pumping unit's pumping unit, N0 points are taken at equal intervals to plot the pumping unit displacement curve, pumping unit load curve, and pumping unit well indicator diagram (in this embodiment, N0 = 256; in other implementations, the value can be chosen according to actual needs). The type of fault occurring during that stroke is determined based on any abnormalities found in the curves, and an automatic alarm is triggered. If the results obtained from the analysis of the curves meet the following criteria, a pumping unit is determined to have experienced a pumping unit pumping unit pumping unit failure (pumping unit pumping unit pumping unit cracking), and an automatic alarm is triggered.

[0051] The specific judgment criteria are as follows:

[0052] ① Displacement curve determination condition: N2 points out of N1 consecutive points in the downstroke of the current donkey head displacement curve do not conform to the normal trend, and the deviation (i.e., the degree of deviation) exceeds 0.5% (i.e. the first set degree threshold), or the deviation of a certain point in the downstroke of the current donkey head displacement curve exceeds 1% (i.e. the second set degree threshold, and the second set degree threshold is greater than the first set degree threshold).

[0053] The deviation is calculated as follows:

[0054]

[0055] In this embodiment, the first set degree threshold is less than the second set degree threshold. N1 > N2 and N2 = 3. The first set degree threshold is 0.5% and the second set degree threshold is 1%. In other implementations, the values ​​of each parameter can be set according to actual needs.

[0056] After a failure of the pumping unit's "donkey head" (a type of welded joint) occurs, during the downward movement of the "donkey head," the load of the rod string inside the wellbore and the weight of the "donkey head" act together on the crack. The elastic deformation as the irregular crack opens is similar to the mechanism of a spring repeatedly expanding and contracting. Due to the generation of elastic deformation, irregular changes in displacement occur, resulting in dense burrs on the downstroke of the pumping unit's "donkey head" displacement curve. In severe cases, the displacement value can change abruptly. Figure 6As shown in the graph, it is clear that the donkey head displacement curve exhibits dense burrs during the downstroke.

[0057] ② Load curve determination condition: During this stroke, the donkey head is at the bottom dead center, i.e. Figure 1 When the crank is rotated to position B2, that is, when the stroke turns from the downstroke to the upstroke, the donkey head is at the bottom dead center. If, at this time, near the bottom dead center, the load value of the current donkey head load curve is lower than the load value of the normal donkey head load curve and decreases by more than 10% (i.e., the third set threshold) for N3 consecutive monitoring points, the load value is lower than the load value of the normal donkey head load curve.

[0058] When a donkey-head crack occurs in a pumping unit, in addition to affecting displacement, the crack's opening and gradual growth at the bottom dead center reversal point causes a significant instantaneous load reduction on the donkey-head, resulting in a depression near the bottom dead center of the complex donkey-head curve. For example... Figure 6 As shown in the graph, it is clear that the donkey head load curve experiences a spring effect at the bottom dead center, resulting in a downward depression.

[0059] ③ Dynamometer determination condition: When, during this stroke, near the bottom dead center, there is one monitoring point on the current pumping unit well dynamometer that shows a load value that is 10% or more lower than the load value of the normal pumping unit well dynamometer (i.e., the fourth set threshold).

[0060] When a pumping unit experiences a "donkey head fracture" failure, the changes in the donkey head displacement and load curves cause corresponding changes in the pumping unit well's dynamometer card. A hanging needle line appears near the bottom dead center of the dynamometer card. When a donkey head fracture is diagnosed, the greater the reduction in load value at this point compared to the normal pumping unit well dynamometer card, and the longer the hanging needle line appears on the image, the larger the fracture is. Figure 7 As shown in the figure, a hanging needle line appears at the bottom dead center of the pumping unit well's dynamometer diagram. The longer the hanging needle line, the stronger the spring effect and the larger the crack in the pumping unit's head.

[0061] In this embodiment, the normal donkey head load curve is the theoretical donkey head load curve of the pumping unit or the donkey head load curve when the pumping unit is running normally, and the normal pumping unit well dynamometer diagram is the theoretical pumping unit well dynamometer diagram or the pumping unit well dynamometer diagram when the pumping unit is running normally.

[0062] In addition, an automatic alarm will be triggered if any of the following phenomena appear in the acquired curves:

[0063] Within the range of crank angle θ from 0 to (π+α), the displacement curve of the pumping unit increases monotonically with time. When the current displacement curve of the current stroke is detected to decrease monotonically or remain unchanged with time within this range, an automatic alarm is triggered.

[0064] Within the range of crank angle θ (π+α) to 2π, the displacement curve of the pumping unit decreases monotonically with time. When it is detected that the current displacement curve of the current stroke is monotonically increasing or unchanged with time when the crank angle is within this range, an automatic alarm is triggered.

[0065] If the minimum value of the current donkey head load curve is detected to be more than 10% lower than the minimum value of the donkey head load curve in a certain stroke within the previous X strokes (i.e., the fifth set threshold), an automatic alarm will be triggered.

[0066] If the minimum value of the current donkey head load curve is detected to be more than 10% lower than the minimum value of the donkey head load curve of a certain stroke in the previous T-hour period (i.e., the sixth set threshold), an automatic alarm will be triggered.

[0067] It should be noted that the various threshold values ​​in this invention were obtained by those skilled in the art through repeated experiments and verifications. For example, in this embodiment, the first threshold value is defined as the percentage of deviation exceeding 0.5%. This threshold value is obtained by comparing the deviation values ​​of multiple consecutive points in the downstroke of the donkey head displacement curve with the deviation values ​​of corresponding consecutive measurement points in the normal donkey head displacement curve of the pumping unit when multiple sets of pumping unit failures with donkey head cracks are obtained by those skilled in the art.

[0068] In summary, the remote diagnosis method for surface faults in a beam pumping unit well of the present invention performs real-time fault diagnosis of surface faults of the pumping unit by using the pumping unit's head displacement curve, head load curve, and pumping unit well dynamometer diagram; and accurately and efficiently identifies the type of fault based on the degree of abnormality of each curve.

Claims

1. A remote diagnostic method for surface faults in a beam pumping unit well, characterized in that, The method includes: Obtain the curves corresponding to the type of pumping unit failure, and determine whether the pumping unit has failed and the type of failure based on all the obtained curves; the curves include only the current donkey head displacement curve and the current pumping unit well dynamometer diagram, or only the current donkey head load curve and the current pumping unit well dynamometer diagram, or the current donkey head displacement curve, the current donkey head load curve and the current pumping unit well dynamometer diagram.

2. The remote diagnosis method for surface faults in beam pumping unit wells according to claim 1, characterized in that, The fault types include donkey head fracture; the curves corresponding to determining whether the pumping unit has a donkey head fracture fault include the current donkey head displacement curve, the current donkey head load curve, and the current pumping unit well dynamometer diagram; if each curve meets all the judgment conditions, then the pumping unit is determined to have a donkey head fracture fault. The determination conditions include: dense spikes or abrupt changes appear in the downstroke of the current donkey head displacement curve; near the transition from the downstroke to the upstroke of the current donkey head load curve, the load values ​​of N3 consecutive points are less than the load values ​​of the corresponding normal donkey head load curve, and the degree of less is greater than a third set threshold; near the transition from the downstroke to the upstroke of the current pumping unit well indicator, the load value of a certain point is less than the load value of the corresponding normal pumping unit well indicator, and the degree of less is greater than a fourth set threshold.

3. The remote diagnosis method for surface faults in beam pumping unit wells according to claim 2, characterized in that, If N2 out of N1 consecutive points in the downstroke of the current donkey head displacement curve have a deviation greater than the first set threshold, then dense burrs are determined to have occurred. The deviation is the ratio of the absolute value obtained by subtracting the measured value of the point from the average displacement of the points before and after the point to the total displacement value of the current stroke, and N1 > N2 ≥ 3.

4. The remote diagnosis method for surface faults in beam pumping unit wells according to claim 2, characterized in that, If the deviation at a certain point in the downstroke of the current donkey head displacement curve is greater than the second set threshold, a sudden change is determined to have occurred; the deviation is the ratio of the absolute value obtained by subtracting the measured value of the point from the average displacement of the points before and after the point to the total displacement value of the current stroke.

5. The remote diagnosis method for surface faults in beam pumping wells according to claim 2, characterized in that, If a donkey-head crack is determined to have occurred in the pumping unit, then when the load value at a certain point is less than the load value of the normal pumping unit well indicator diagram, the greater the degree to which the value is less than the fourth set degree threshold, the larger the crack that has caused the donkey-head crack.

6. The remote diagnostic method for surface faults in beam pumping unit wells according to any one of claims 1 to 5, characterized in that, If the displacement of the donkey head during the upward stroke of the current displacement curve decreases or remains unchanged over time, an alarm will be issued.

7. The remote diagnostic method for surface faults in beam pumping wells according to any one of claims 1 to 5, characterized in that, If the displacement of the donkey head during the downstroke of the current displacement curve increases or remains unchanged over time, an alarm will be issued.

8. The remote diagnosis method for surface faults in beam pumping wells according to claim 1, characterized in that, If the minimum value of the current donkey head load curve is less than the minimum value of the donkey head load curve of a certain stroke within the previous X strokes, and the degree of less than is greater than the fifth set threshold, then an alarm will be triggered.

9. The remote diagnosis method for surface faults in beam pumping wells according to claim 1, characterized in that, If the minimum value of the current donkey head load curve is less than the minimum value of a certain donkey head load curve in the previous T time periods, and the degree of less than is greater than the sixth set threshold, then an alarm will be triggered.

10. The remote diagnosis method for surface faults in a beam pumping unit well according to claim 2, characterized in that, The normal donkey head load curve is the theoretical donkey head load curve of the pumping unit or the donkey head load curve when the pumping unit is operating normally; the normal pumping unit well dynamometer diagram is the theoretical pumping unit well dynamometer diagram or the pumping unit well dynamometer diagram when the pumping unit is operating normally.

Images